Techniques for compressing differential samples of bandwidth-limited data to reduce bandwidth and power consumption by an interface转让专利
申请号 : US11970887
文献号 : US07835435B2
文献日 : 2010-11-16
发明人 : Samir J. Soni , Pravin Pramakanthan , Clive K. Tang , Bing Xu
申请人 : Samir J. Soni , Pravin Pramakanthan , Clive K. Tang , Bing Xu
摘要 :
权利要求 :
What is claimed is:
说明书 :
Embodiments of the subject matter described herein relate generally to data communications, and more particularly, to communication of bandwidth-limited data over a serial interface between a data source integrated circuit (IC) and a data sink IC.
When a real-time, digitized, bandwidth-limited signal is transmitted over a serial interface, it is desirable to improve operational speed of the interface and to reduce power consumption. This is especially true as bandwidth requirements increase.
Thus, it is desirable to provide techniques for high-speed data transfer over a serial interface without loss of information while reducing bandwidth utilization and power consumption of the serial interface. Furthermore, other desirable features and characteristics will become apparent from the subsequent detailed description and the appended claims, taken in conjunction with the accompanying drawings and the foregoing technical field and background.
A more complete understanding of the subject matter may be derived by referring to the detailed description and claims when considered in conjunction with the following figures, wherein like reference numbers refer to similar elements throughout the figures.
The following detailed description is merely exemplary in nature and is not intended to limit the invention or the application and uses of the invention. The word “exemplary” is used herein to mean “serving as an example, instance, or illustration.” Any embodiment described herein as “exemplary” is not necessarily to be construed as preferred or advantageous over other embodiments. All of the implementations described below are exemplary implementations provided to enable persons skilled in the art to make or use the invention and are not intended to limit the scope of the invention which is defined by the claims. Furthermore, there is no intention to be bound by any expressed or implied theory presented in the preceding technical field, background, brief summary or the following detailed description.
Techniques and technologies may be described herein in terms of functional and/or logical block components and various processing steps. It should be appreciated that such block components may be realized by any number of hardware, software, and/or firmware components configured to perform the specified functions. For example, an embodiment of a system or a component may employ various integrated circuit components, e.g., memory elements, digital signal processing elements, logic elements, look-up tables, or the like, which may carry out a variety of functions under the control of one or more microprocessors or other control devices. In addition, those skilled in the art will appreciate that embodiments described herein are merely exemplary implementations.
The connecting lines shown in the various figures contained herein are intended to represent example functional relationships and/or physical couplings between the various blocks, circuits, components, devices, elements or modules. It should be noted that many alternative or additional functional relationships or physical connections may be present in other embodiments of the subject matter. The following description refers to blocks, circuits, components, devices, elements or modules being “connected” or “coupled” together. As used herein, unless expressly stated otherwise, “connected” means that one block, circuit, component, device, element or module is directly joined to (or directly communicates with) another block, circuit, component, device, element or module, and not necessarily mechanically. Likewise, unless expressly stated otherwise, “coupled” means that one block, circuit, component, device, element or module is directly or indirectly joined to (or directly or indirectly communicates with) another block, circuit, component, device, element or module, and not necessarily mechanically. Thus, although the schematics illustrated in
Terminology
As used herein, the term “bandwidth-limited data stream or signal” refers to a stream of data samples wherein the frequency content carried by the data stream is finite with respect to data rate. For instance, in one implementation, the Fast Fourier Transform (FFT) of a data stream samples will result in bandwidth limited response. The terms “bandwidth-limited data stream” and “bandwidth-limited data signal” are used interchangeably throughout this description.
As used herein, the term “frame” refers to a unit of data comprising either: a default bit code and a number (p) of raw data samples, or a particular compression bit code and a number (p) of compressed-differential samples of the raw data. The terms “frame” “packet” and “block” are used interchangeably throughout this description.
As used herein, the term “mode of operation” refers to a specific configuration of a data source IC and a data sink IC. The “mode of operation” can determine the data rate, data bandwidth and other system level requirements, such as the configuration and protocols the data source IC and data sink IC agree upon for data exchange. In some implementations, the modes of operation can include, for example, GSM/EDGE, WCDMA, LTE, 4G, etc. In a given mode of operation, values of variables, such as the number (p) of samples, the number (i) of differentiation modules, threshold values (threshold_1 . . . threshold_T), are predetermined based on the mode of operation.
As used herein, the term “representable” can be interpreted as meaning “capable of being represented” and/or “can be represented by.” For instance, one exemplary usage of this term is that, a sample is “representable” using a number of bits.
Overview
In the exemplary embodiments which will now be described, techniques and technologies will be described for compressing differential samples of bandwidth-limited data and coding the compressed differential samples to reduce bandwidth and power consumption when communicating bandwidth-limited data over a serial interface which couples one integrated circuit to another integrated circuit. However, it will be appreciated by those skilled in the art that the same or similar techniques and technologies can be applied to other types of interfaces.
The communication system 10 comprises a data source IC 15, a serial interface 55, and a data sink IC 25. In this embodiment, the serial interface 55 includes the interface modules 50, 92 which enable communication between the data source IC 15 and the data sink IC 25. The data source IC 15 includes the interface module 50 and a differential compression coding (DCC) module 20 which receives a data stream 5 of n raw data samples and then generates, based on the bandwidth-limited data 5, a frame 85/90 comprising either (a) a frame 85 of the actual or “raw” samples of bandwidth-limited data 5 or (b) a frame 90 of compressed, differential samples of the bandwidth-limited data 5. In one implementation, the frame 85 comprises a default bit code (bcd) and a number (p) of consecutive raw data samples of the raw data samples, where p≦n. In one implementation, the frame 90 comprises a particular or specific compression bit code (bct) and a number (p) of consecutive compressed-differential samples which are generated based on at least some of the some of the raw data samples, where p≦n.
The interface module 50 of the data source IC 15 communicates the frame 85 or 95 to the interface module 92 of data sink IC 25. The data sink IC 25 also includes an integration and decompression (ID) module 95. As will be described in greater detail below, upon receiving the frames 85/90 from the data source IC 15, the ID module 95 processes frames 85/90 to regenerate or extract the raw data samples of the bandwidth-limited data 5. For example, when the ID module 95 receives a frame 90, the ID module 95 performs decompression operations to reproduce the correct number of bits used to represent each of the differential samples, and then performs integration operations on the differential samples of the frame 90 to generate the actual samples of the original bandwidth-limited data 5.
Before describing processing performed by the DCC module 20 and the ID module 95 in more detail, an exemplary IC architecture will be now described below with reference to
The disclosed techniques and technologies can be utilized for transmitting bandwidth-limited data over any serial interface including those used in conjunction with wireless communication standards such as GSM/EDGE, WCDMA, LTE, 4G, WiMAX, etc. One such serial interface is defined by the DigRF standard. In this implementation, the AS interface 155 includes DigRF interface hardware modules 150, 192 for communicating data between the RFIC 115 and the BBIC 125. In this implementation, the DCC module 120 can be implemented as part of or in conjunction with a high-speed asynchronous serial interface. In this implementation, the interface 155 is described as being compliant with the DigRF physical interface standard, which defines a digital serial interface that is implemented between a baseband IC (BBIC) 115 and a radio frequency IC (RFIC) 125 used in wireless communication devices. The DigRF standard is being developed by the DigRF working group under MIPI alliances. The DigRF standard places few constraints on the internal architectures of the ICs, and allows for any combination of supporting chip sets, and supports a variety of third Generation Partnership Project (3GPP) and third Generation Partnership Project 2 (3GPP2) standards. 3GPP-based standards such as GSM, EDGE, CDMA, WCDMA, LTE, UMTS, WiMax, IEEE 802.16, IEEE 802.20 and IEEE 802.11a/b/g. The asynchronous, digital serial interface specified by the DigRF standard replaces analog interfaces used in previous generations of wireless communication device architectures and eliminates the need for intermediate mixed signal devices. The DigRF standard defines an efficient physical interconnection between the RFIC and BBIC which allows the RFIC and the BBIC to communicate directly with each other. Among other things, the DigRF standard describes or will describe the logical, electrical and timing characteristics of the digital interface with sufficient detail to allow physical implementation of the interface, and with sufficient rigor that implementations of the interface from different suppliers are fully compatible at the physical level.
When the antenna 45 receives a complex radio frequency (RF) signal, the antenna 45 passes the complex RF signal to the front end module 135, which in turn communicates the RF signal to a demodulator 116. The complex RF signal can sometimes be referred to as an incoming or “downlink” RF signal. The demodulator 116 demodulates the complex RF signal and outputs analog I channel signals to one of the analog-to-digital converters 117 and analog Q channel signals to the other one of the analog-to-digital converters 117. Based on the analog I and Q channel signals, one of the analog-to-digital converter 117 outputs a digital I channel signal, and the other analog-to-digital converter 117 outputs a digital Q channel signal, respectively.
The bandwidth-limiting filter 118 receives the digital I/Q channel signals and filters the digital I/Q channel signals to limit the bandwidth of the digital I/Q channel signals The bandwidth-limiting filter 118 outputs digital samples 105 of bandwidth-limited I channel data, and samples of bandwidth-limited Q channel data. After extracting I channel data and Q channel data samples 105, referred to herein as raw, bandwidth limited I/Q data samples, from the incoming radio frequency (RF) signal, the bandwidth-limiting filter 118 communicates the data stream of raw, bandwidth-limited I/Q data samples 105 to the DCC module 120.
As will be described below, for example, with reference to
Block 205 illustrates a data stream 5 of raw, bandwidth-limited data samples (1 . . . n). The individual samples of the bandwidth-limited data stream 5 each require a certain number of bits to represent the value of those samples. As such, any of the raw data samples requires, at most, some maximum number (b) of bits to represent its value. In other words, “b” is total number of bits required to represent each raw data sample of the data stream 5.
When the DCC module 20 receives samples of the bandwidth-limited data 5, at step 230, the DCC module 20 performs one or more differentiation operations on each of the raw data samples to generate or compute “differential samples” based on the raw, bandwidth-limited data samples. This results in differential data stream of ith order differential samples.
At step 245, the DCC module 20 selects a number (p) of ith order differential samples of the differential data stream for further processing. Here, p represents a number of consecutive raw data samples selected for further processing, where p≦n.
At step 250, the DCC module 20 then determines whether the number (p) of ith order differential samples can be represented using a fewer number of bits (b−bt) than the maximum number (b) of bits required to represent corresponding samples of the bandwidth-limited data 5.
When the number (p) of ith order differential samples can be represented using a fewer number of bits (b−bt) than the maximum number (b) of bits required to represent corresponding raw data samples of the bandwidth-limited data 5, then at step 260, the DCC module 20 compresses or “trims” each of the number (p) of ith order differential samples by removing the number of bits (bt) from each differential sample to produce a compressed representation of the number (p) of ith order differential samples. At step 275, the DCC module 20 generates a specific compression bit code (bct) which indicates the number of bits (bs) to be removed or trimmed from each differential sample. At step 290, the DCC module 20 generates a compressed frame 90 based on the compressed representation of the differential samples and attaches the specific compression bit code (bct) to the frame 90.
By contrast, when the one or more of ith order differential samples can not be represented using a fewer number of bits (b−bt) than the maximum number (b) of bits required to represent corresponding raw data samples of the bandwidth-limited data 5, then at step 280, the DCC module 20 generates a default bit code (bcd) which indicates that no bits are removed or trimmed from the raw data samples of the bandwidth-limited data 5. At step 285, the DCC module 20 generates a frame 85 using the raw data samples of the bandwidth-limited data 5 and attaches the default bit code (bcd) as part of the frame 85.
Differential Computation
The differential computation module 304 receives n raw data samples (1 . . . n) of a bandwidth-limited data stream 305 from a data source, such as the data source IC 15 of
Based on the n raw data samples (1 . . . n) of the bandwidth-limited data stream 305, the differential computation module 304 computes the first . . . ith order differential samples corresponding to the raw data samples of the bandwidth-limited data stream 305 to generate a differential data stream 345 of n ith order differential samples. An example of the differential data stream 345 is illustrated in
The differential computation module 304 includes at least one differentiator module (e.g., when i is equal to 1), but more than one differentiator module or multiple differentiator modules (e.g., i≧1) can be implemented in other implementations which depend, for example, on the mode of operation and characteristics of the bandwidth-limited data stream 305 (e.g., the sampling rate to bandwidth relationship of the bandwidth-limited data stream 305). In this exemplary implementation, the differential computation module 304 comprises a first differentiator module 310 through an ith differentiator module 340, where i≧1.
The first differentiator module 310 receives the raw data samples (1 . . . n) of the bandwidth-limited data stream 305, and generates or computes a first order differential data stream 315 of first order differential samples by computing a first order differentials corresponding to each of the raw data samples (1 . . . n) of the data stream 305. For instance, with respect to a raw data sample (Ij+1) the first differentiator module 310 can calculate a first order differential sample d1Ij+1=Ij+1−Ij, where Ij+1 represents a raw data sample of the bandwidth-limited data stream 305, Ij represents a prior raw data sample of the bandwidth-limited data stream 305, and d1Ij+1 represents the first order differential sample. The first differentiator module 310 communicates each of the first order differential samples 315 to the ith differentiator module 340.
The ith differentiator module 340 receives the first order differential data stream 315 of first differential samples (or alternatively (i−1)th order differential data stream 315 of (i−1)th differential samples if more differentiator modules are implemented between first differentiator module 310 and the ith differentiator module 340), and generates or computes an ith order differential data stream 345 of ith order differential samples by computing ith order differentials corresponding to each of the first differential samples. For instance, the ith differentiator module 340 can calculate an ith order differential sample diIj+1=d(i−1)Ij+1−d(i−1)Ij, where diIj+1 represents an ith order differential sample, where d(i−1)Ij represents a prior differential sample, and where d(i−1)Ij+1 represents the ith order differential sample.
As described below, for a given amount of raw data samples of the bandwidth-limited data stream 305, the distribution of values of the ith order differential samples 345 is more “Gaussian” than the distribution of values of the raw data samples of the bandwidth-limited data stream 305, the first order differential samples 315 of the bandwidth-limited data signal (or the (i−1)th differential samples when more than two differentiator modules are implemented). As used herein, and as will be described in more detail below, the term “Gaussian,” when used in conjunction with the term distribution, refers to an approximately normal distribution which has a probability density similar to the shape of a bell.
Raw data samples of the bandwidth-limited data stream 305 have values which range between a maximum value (+MAX) and a minimum value (−MAX). The variable b represents the number (b) of bits are required to represent the maximum value of any raw data sample that are within the range 2MAX. In other words, the variable b is the maximum number of bits required to represent any raw data sample of the bandwidth-limited data stream 305. Because the data stream 305 is “bandwidth-limited,” the likelihood that values of the raw data samples will vary outside the range (i.e., between the maximum value (+MAX) and the minimum value (−MAX)) from one sample to the next is relatively small. Because the difference between any two consecutive ith order differential samples 345 is smaller than the difference between any two consecutive first order differential samples 315, a lesser number of bits (b−bt) are required to represent the ith order differential samples 345. For instance, in comparison to (i−1)st order differential samples (not illustrated in
Thus, in some scenarios, the ith order differential samples 345 can be represented using with lesser number of bits (b−bt) in comparison to the raw data samples of the bandwidth-limited data stream 305, the first order differential samples of the bandwidth-limited data stream 305, the (i−1)th differential of those samples, etc.
Dynamic Compression and Coding
By generating the ith order differential samples 345, the dynamic compression and coding module 346 can reduce the number of bits required to represent the respective values of most of the raw data samples 305 so that the number of bits required to represent each of the differential samples 345 can subsequently be reduced by a number (bt) of bits.
The dynamic compression and coding module 346 applies dynamic compression techniques to reduce or trim the number (b) of bits required to represent each raw sample 305 to generate compressed-differential samples 347′ of the raw data samples (1 . . . n). The dynamic compression and coding module 346 also applies the dynamic coding techniques to generate additional bit coding (bc) information to be included along with the compressed-differential samples 347′ such that the compressed-differential samples 347′ can be decompressed at the receiver.
In the exemplary implementation illustrated in
The dynamic compression and coding module 346 receives the data stream 305 of raw, bandwidth-limited data samples (1 . . . n) and the differential data stream 345 of ith order differential samples which are generated by the differential computation module 304 based on the raw, bandwidth-limited data samples (1 . . . n).
The threshold comparator module 350 comprises a module 352 for designating a number (p) of consecutive ith order differential samples 345 for comparison to various threshold values (threshold_1 . . . threshold_T). The exemplary embodiment illustrated in
The threshold comparator module 350 includes a module 352 for designating a number (p) of consecutive ith order differential samples 345 to be compared against each of the threshold values (threshold—1 . . . threshold_T). The value p can be predetermined based on the mode of operation (e.g., mode of operation could be GSM/EDGE, WCDMA, LTE, 4G etc). The value of p also depends on the probability of having one larger value within so many consecutive samples. These values can be determined by simulations or other means.
The threshold comparator module 350 compares each of the p consecutive ith order differential samples 345 against each of a plurality of threshold values (threshold—1 . . . threshold_T) to determine if each of the p consecutive ith order differential samples 345 can be fit within a certain one of the threshold values (threshold—1 . . . threshold_T) without any loss. This way, the threshold comparator module 350 can dynamically determine whether bit savings can be achieved by compressing the p consecutive ith order differential samples 345 thereby reducing overhead.
If the threshold comparator module 350 determines that compressing the p consecutive ith order differential samples 345 is possible (and hence that bit savings can be increased), the threshold comparator module 350 communicates the threshold comparison output to the bit code generator module 370, which generates a specific compression bit code (bct) 375 for threshold t, where t is the smallest threshold value (out of the threshold—1 , threshold—2 , . . . threshold_T) for which all of the p consecutive ith order differential samples 345′ are smaller than threshold t. The specific compression bit code (bct) 375 indicates that the p consecutive ith order differential samples 345′ can be compressed. The threshold comparator module 350 communicates the specific compression bit code (bct) 375 to the sample selector and bit trimming module 360 and the framing module 382.
As described below, the compression module 364 selects the p consecutive ith order differential samples 345′ and determines how many bits are to be removed or “trimmed” from each of the p consecutive ith order differential samples 345′ to generate the number (p) of consecutive compressed-differential samples 347′.
If the threshold comparator module 350 determines that compressing the p consecutive ith order differential samples 345 is not possible (and hence that bit savings can not be increased by compression of the samples 345), the bit code generator module 370 generates a default bit code (bcd) 380. In other words, when the threshold comparator module 350 determines that one or more of the p consecutive ith order differential samples 345′ are not smaller than any of the threshold values (threshold—1 . . . threshold_T), the threshold comparator module 350 generates the default bit code (bcd) 380 which indicates that the p consecutive ith order differential samples 345′ can not be compressed and therefore bit savings can not be increased by compressing the p consecutive ith order differential samples 345′. The threshold comparator module 350 communicates the default bit code (bcd) 380 to the sample selector and bit trimming module 360 and the framing module 382. As described below, in response to the threshold comparison output, the compression module 364 selects the number (p) of consecutive raw data samples 305′ and does not perform any compression.
As described above, the bit code generator module 370 generates an appropriate bit code 375, 380 based on the result of the threshold comparison output, 355. In one scenario, a specific compression bit code (bct) 375 for threshold t is generated, and in another scenario, a default bit code (bcd) 380 is generated which indicates “raw data” is selected and that no compression is to be performed on the number (p) of consecutive raw data samples 305′. For example, if all of the p of consecutive ith order differential samples 362 are smaller than a particular threshold values (threshold—1 . . . threshold_T), the bit code generator module 370 generates an appropriate specific compression bit code (bct) 375 which indicates the smallest threshold t (out of threshold—1 , threshold—2 , . . . threshold_T) for which all of the p consecutive ith order differential samples 345′ are smaller than threshold t). The value of the specific compression bit code (bct) 375 for threshold t depends on the number of threshold values used by the threshold comparator module 350, and for a selected number of different threshold values there will be 2bc possible bit codes, where bc is log2 (T+1) bits required to represent all possible bc codes. On the other hand, if any of the p consecutive ith order differential samples 345′ are greater than or equal to the largest threshold value amongst the selected threshold values (threshold—1 . . . threshold_T) used by the threshold comparator module 350 (i.e., if any of the p samples can not be fit within the threshold values (threshold—1 . . . threshold_T)), the bit code generator module 370 generates the default bit code (bcd) 380 that indicates that no bit savings are possible.
The sample selector module 363 selects either: (a) the p consecutive ith order differential samples 345′ of the raw data samples (1 . . . n) for transmission (instead of the number (p) of consecutive raw data samples 305′), or (b) the number (p) of consecutive raw data samples 305′ (as illustrated in
Based on the threshold comparison output, the sample selector module 363 designates either: a number (p) of consecutive ith order differential samples 345′ which correspond to p consecutive raw data samples of the bandwidth-limited data stream 305, or a number (p) of consecutive raw data samples 305′ for transmission. When the p consecutive ith order differential samples 345′ are selected, the compression module 364 performs bit trimming on the number (p) of consecutive ith order differential samples 345′. The compression module 364 outputs either the selected one of the number (p) of consecutive compressed-differential samples 347′ or the number (p) of the raw data samples (1 . . . n) 305′.
When the selector module 363 selects the p consecutive ith order differential samples 345′, the compression module 364 performs bit trimming on the number (p) of consecutive ith order differential samples 345′ to generate a number (p) of consecutive compressed-differential samples 347′. An example of the p consecutive compressed-differential samples 347′ is illustrated in
Depending on whether the specific compression bit code (bct) 375 or default bit code (bcd) 380 is provided to the framing module 382, the framing module 382 generates either a frame 385 comprising default bit code (bcd) 380 and a number (p) of consecutive raw data samples 305′, or a frame 390 comprising a particular or specific compression bit code (bct) 370 and the number (p) of consecutive compressed-differential samples 347′.
For example, when the specific compression bit code (bct) 375 for threshold t is provided to the framing module 382, the framing module 382 generates the frame 390. As illustrated in
When a data sink IC (e.g., data sink 25 of
When the data sink IC receives the frame 385, the data sink IC uses the default bit code (bcd) 380 to determine that the frame 385 has not been compressed.
In the method, block 605 represents a bandwidth-limited data stream 605 which comprises n raw data samples (1 . . . n).
At step 640, based on the n raw data samples (1 . . . n) of the bandwidth-limited data stream 305, first . . . ith order differential samples corresponding to the n raw data samples (1 . . . n) are computed to generate a differential data stream 345 of n ith order differential samples.
At step 645, a number (p) of consecutive ith order differential samples 345 are selected for comparison to various threshold values (threshold—1 . . . threshold_T), where p≦n. The value p can be predetermined based on the mode of operation (e.g., mode of operation could be GSM/EDGE, WCDMA, LTE, 4G etc). The value of p also depends on the probability of having one larger value within so many consecutive samples. These values can be determined by simulations or other means. Although not shown, a number of threshold values (threshold—1 . . . threshold_T) have been specified or selected depending on the mode of operation. As described below, the plurality of threshold values (threshold—1 . . . threshold_T) can be selected such that when the values of each of the p ith order differential samples are less than a particular one of the threshold values (threshold—1 . . . threshold_T), the number (p) of consecutive ith order differential samples 345 can be represented with b−bt bits, where b is the number of bits used to represent the value of any of the raw data samples (1 . . . p) and bt is an integer number of bits to be trimmed from each of the p raw data samples (1 . . . p).
At step 650, it is determined whether the number of bits required to represent each of the p ith order differential samples 345 can be subsequently reduced by a number (bt) of bits with respect to the number of bits required to represent the corresponding values of each of the p raw data samples 305. In other words, step 650 dynamically determines whether bit savings can be achieved by compressing the p consecutive ith order differential samples 345 thereby reducing overhead. In this implementation, step 650 will be described with reference to steps 652-658. At step 652, the number (p) of consecutive ith order differential samples 345 are compared to a smallest threshold value (threshold t) of the various threshold values (threshold—1 . . . threshold_T).
At step 652, each of the p consecutive ith order differential samples 345 are compared against the smallest one of the plurality of threshold values (threshold—1 . . . threshold_T) to determine if each of the p consecutive ith order differential samples 345 can be fit within a certain one of the threshold values (threshold—1 . . . threshold_T) without any loss. At step 656, it is determined whether each of the p consecutive ith order differential samples 345 is less than a current threshold value, which in the first iteration of step 656 is the smallest threshold value, threshold—1 . If one or more of the p consecutive ith order differential samples 345 is greater than or equal to the current threshold value (threshold—1 ), then method 600 proceeds to step 658, where it is determined whether the current threshold value (threshold—1 ) used at step 656 is the last threshold value (threshold_T).
If the current threshold value (threshold—1 ) used at step 656 is not the last threshold value (threshold_T), then the method 600 proceeds to step 659, where the number (p) of consecutive ith order differential samples 345 are compared to the next threshold value (threshold—2 ) of the various threshold values (threshold—1 . . . threshold_T), and the method 600 returns to step 656, where it is determined whether each of the p consecutive ith order differential samples 345 is less than a current threshold value (threshold—2 ), which in the second iteration of step 656 is threshold—2 . Steps 656, 658 and 659 continue to iterate until the values of the p consecutive ith order differential samples 345 have been compared against the last threshold value (threshold_T), and if the p consecutive ith order differential samples 345 do not fit within any of the threshold values (threshold—1 . . . threshold_T), then the method 600 continues to step 690.
If one or more of the p consecutive ith order differential samples 345′ are greater than or equal to each of the particular threshold values (threshold—1 . . . threshold_T) (i.e., one or more of the p consecutive ith order differential samples 345′ are not smaller than any of the threshold values (threshold—1 . . . threshold_T)), then the p consecutive ith order differential samples 345′ do not fit within at least one of the threshold values (threshold—1 . . . threshold_T), and therefore compression of the p consecutive ith order differential samples 345′ is not possible. Thus, if the current threshold value (threshold—1 ) used at step 656 is the last threshold value (threshold_T), then the method 600 proceeds to step 690.
At step 690, a default bit code (bcd) is generated and a number (p) of consecutive raw data samples 305′ are selected for framing and transmission, where p≦n. The value p can be predetermined based on the mode of operation, for example, by simulations. The default bit code (bcd) 380 indicates that bit savings can not be increased by compressing the p consecutive ith order differential samples 345′, and that the p consecutive ith order differential samples 345′ can not be compressed. As such, the number (p) of consecutive raw data samples 305′ are selected and no compression is performed.
At step 695, a regular frame 385 is created which includes the default bit code (bcd) and the number (p) of consecutive raw data samples. The default bit code (bcd) 380 indicates “raw data” is present in the frame 385 and that no compression has been performed on the number (p) of consecutive raw data samples 305′.
At step 697, the regular frame 385, which includes the default bit code (bcd) and the number (p) of consecutive raw data sample 305′, is transmitted to a data sink IC. When a data sink IC receives the frame 385, the data sink IC uses the default bit code (bcd) 380 to determine that the frame 385 has not been compressed.
If each of the p consecutive ith order differential samples 345′ is less than the current threshold value used at step 656, then it is determined that the p consecutive ith order differential samples 345′ can be compressed (and hence that bit savings can be increased), and method 600 proceeds to step 660.
At step 660, a specific compression bit code (bct) is generated, and the p consecutive ith order differential samples 345′ are selected for framing and transmission instead of the number (p) of consecutive raw data samples 305′. The specific compression bit code (bct) indicates that the p consecutive ith order differential samples 345′ can be compressed. The specific compression bit code (bct) is generated for threshold t, where t is the smallest threshold value (out of the threshold—1 , threshold—2 , . . . threshold_T) for which all of the p consecutive ith order differential samples 345′ are smaller than threshold t. The specific compression bit code (bct) 375 indicates the smallest threshold t (out of threshold—1 threshold—2 , . . . threshold_T) for which all of the p consecutive ith order differential samples 345′ are smaller than threshold t. The value of the specific compression bit code (bct) 375 for threshold t depends on the number of threshold values used by the threshold comparator module 350, and for a selected number of different threshold values there will be 2bc possible bit codes, where bc is log2 (T+1) bits required to represent all possible bc codes.
At step 670, dynamic compression techniques are applied to the p consecutive ith order differential samples 345′. Based on the specific compression bit code (bct), it is determined how many bits (bt) are to be removed or “trimmed” from each of the p consecutive ith order differential samples 345′, and a number of bits (bt) are then removed or “trimmed” from each of the p consecutive ith order differential samples 345′ to compress each of the p consecutive ith order differential samples 345′ and generate the number (p) of consecutive compressed-differential samples 347′. In one implementation, a number of most significant bits (MSBs) are trimmed or removed from each of the p consecutive ith order differential samples 345′ to generate a number (p) of consecutive compressed-differential samples 347′. The number (b) of bits required to represent each of the compressed-differential samples 347′ is reduced in comparison to the number of bits required to represent each of the p consecutive ith order differential samples 345′. The number (p) of consecutive compressed-differential samples 347′ includes only b−bt least significant bits (LSBs) for each of the p values. The value of each of consecutive compressed-differential samples 347′ is represented using (b−bt) total bits.
At step 680, a compressed frame 390 is generated which includes the specific compression bit code (bct) 375 for threshold t and for the amount of bit trimming bt applied for the number (p) of consecutive compressed-differential samples 347′.
At step 685, the compressed frame 390 is transmitted. The receiver of compressed frame 390 uses the specific compression bit code (bct) 375 to decompress the number (p) of consecutive compressed-differential samples 347′.
Exemplary Simulation of Differential Computation
To illustrate the concept of differential computation and the results of a differential computation, an example will now be described with reference to FIGS. 7 and 8A-8E.
The differential computation module 704 comprises four differentiator modules 710-740 including a first differentiator module 710, a second differentiator module 720, a third differentiator module 730 and an ith differentiator module 740, where the variable (i) is equal to four (4), and therefore, the ith differentiator module 740 will be described below as a fourth differentiator module 740. In the particular implementation described below, in which the variable (i) is equal to four (4), the differential computation module 704 takes the first, second, third . . . ith order differential samples corresponding to the raw I/Q data samples of the bandwidth-limited GSM/EDGE data stream 705, and generates first, second, third . . . ith order differential samples 715, 725, 735, 745 of the raw I/Q data samples of the bandwidth-limited GSM/EDGE data stream 705. In other implementations, the variable (i) can be greater than four (4) in which case there may be more than four differentiator modules 710-740 in total, while in other implementations, the variable (i) can be less than four (4) in which case there may be fewer than four differentiator modules 710-740 in total.
In the following simulation, the differential computation module 704 receives the raw I/Q data samples of the bandwidth-limited GSM/EDGE data stream 705 from a data source (e.g., extracted from a complex RF signal, such as those from the band-limiting filter 118 of
The first differentiator module 710 generates a first order differential data 715 representation of each of the raw I/Q data samples of the bandwidth-limited I/Q data stream 705, which is communicated to the second differentiator module 720. For instance, the first differentiator module 710 computes the first order differential of each of the raw I/Q data samples of the data stream 705 by calculating d1Ij+1=Ij+1−Ij to generate a first order differential data 715 representation of each of the raw I/Q data samples of the bandwidth-limited I/Q data stream 705.
The second differentiator module 720 generates second order differential data 725 representation of the raw I/Q data samples of the bandwidth-limited I/Q data stream 705 based on the first order differential data 715, and communicates the second order differential data 725 to the third differentiator module 730.
The third differentiator module 730 generates a third order differential data 735 representation of the raw I/Q data samples of the bandwidth-limited I/Q data stream 705 based on the second order differential data 725, and communicates the third order differential data to the fourth differentiator module 740.
The fourth differentiator module 740 generates a fourth order differential data 745 representation of the raw I/Q data samples of the bandwidth-limited I/Q data stream 705 based on the third order differential data 735.
Exemplary Bit-Savings Achieved with Differential Dynamic Compression and Coding
Based on the simulations, ran over millions of I/Q samples, it becomes clear that the percentage of bit savings can vary depending on variables which include: the number (i) of differentiation modules, the number (p) of consecutive ith order differential samples which are selected, the number of threshold values (threshold—1 . . . threshold_T) that the p consecutive ith order differential samples are compared against, and the specific threshold values (threshold—1 . . . threshold_T) that are used for the comparison. Although not illustrated in
For example, in the WCDMA mode of operation when i is equal to 4 (i=4), the number (p) of consecutive ith order differential samples is one (1), and the one (1) ith order differential sample is compared against threshold values of ½, ¼ and 1/64, the percentage of bit savings is 1.67%. By contrast, when the number (p) of consecutive ith order differential samples is eight (8), and the eight (8) ith order differential samples are compared against the same threshold values, the percentage of bit savings is 8.64%.
Similarly, in the GSM/EDGE mode of operation when i is equal to 4 (i=4), the number (p) of consecutive ith order differential samples is two (2), and the two (2) ith order differential samples are compared against threshold values of ½, ¼, ⅛, 1/16, 1/32 and 1/64, the percentage of bit savings is 3.92%, but when the number (p) of consecutive ith order differential samples is increased to four (4), and the four (4) ith order differential samples are compared against threshold values of ½, ¼ and 1/64, the percentage of bit savings is 9.95%.
Another factor that influences the percentage of bit savings is the nature of data being compressed. The bandwidth limiting performed on the data, and its sampling rate, for example, may also affect the percentage of bit savings.
Thus, by adding a negligible amount of information, significantly lesser numbers of bits are transmitted as compared to sending the raw bandwidth-limited data. Appropriate values for the different variables described above can be selected and/or predetermined to help reduce and/or minimize the total number of bits that are transferred over an interface and to reduce the overhead associated with code bits. By reducing the total number of bits being transferred, bandwidth is saved. In addition, current consumption can also be reduced because the total current consumption of pads, line driver, line receiver and all other supporting circuitry of a serial interface is directly proportional to the number of bits transferred.
While at least one example embodiment has been presented in the foregoing detailed description, it should be appreciated that a vast number of variations exist. It should also be appreciated that the example embodiment or embodiments described herein are not intended to limit the scope, applicability, or configuration of the claimed subject matter in any way. Rather, the foregoing detailed description will provide those skilled in the art with a convenient roadmap for implementing the described embodiment or embodiments. It should also be understood that various changes can be made in the function and arrangement of elements without departing from the scope defined by the claims, which includes known equivalents and foreseeable equivalents at the time of filing this patent application.